Adobe After Effects - How do I maintain best resolution when

Author

How do I maintain best resolution when

Mark_Silvestri@adobeforums.com

2006-12-06, 8:15 pm

I have a problem that must have a simple solution.

I have scaled down and masked some video footage to fit on the body of a musical note. I go on to transform this musical note in 3D space. However, when those transformations involve decreasing the z-position of the camera or increasing the z-position of
the null to which the camera is parented (effectively moving closer to the musical note, in either case), that scaled-down footage is blurred - I seem to be stuck with the scaled-down resolution of that footage instead of the best resolution of that foota
ge for its size at any point in the comp. It seems there should be some kind of switch (similar to the "continuously rasterize" switch for vector art) that would resolve this.

My current solution is a clunky one. I leave the footage scaled to 100% and blow up my vectored musical note to achieve the right proportions. However, since I often use only a portion of the video frame and not the entire frame size, the scale percentage
of my musical note is different for each note. This is difficult because I now have a different-sized note for each of the many notes I use in my comp. This dramatically complicates my transformations for each note.

I would prefer to keep my notes the same size then simply scale each note's footage (and adjust anchor point) while maintaining best resolution in all transformations. I am using AE Pro 7.0. Is this possible?

Aaron_Cobb@adobeforums.com

2006-12-06, 8:15 pm

First things first, if you are masking your footage to fit on the musical note, it might be worthwhile for you to read up on Track Mattes. They let you use the alpha or luminance of one layer to matte another layer, and save yourself some work drawing mas
ks to match (if that is what you have done in this instance).

As for your resolution problems, it all depends on just how you have your comp set up. It sounds like you aren't nesting comps here -- is that right? If the footage isn't nested, then it should be rendered at the native resolution of the comp, and there a
re no additional transformations to collapse. If the footage is coming out blurry, it could be the result of separating fields. You lose half your vertical resolution when you separate fields, and there isn't really any good way to get around this. If the
source footage is interlaced you have to separate fields when you move the layer (and even if you don't move the layer, you'd better separate fields or really know what you are doing).

It could also be that you've got your comp set to preview at less than full resolution, or you've got depth of field enabled on your camera, or you've got some funky OpenGL things going on.

If, on the other hand, I'm misunderstanding and you are nesting your comps, then you should be aware that every nested comp gets rendered at its own native resolution before being composited into the comp that contains it. If you scale down layers in the
nested comp, then scale up the nested comp to compensate, you will lose resolution unless you collapse transformations, which forces the contents of the nested comp to be rendered as if they lived in the main comp. But collapsing transformations changes t
he render order, so the result will look different in other, possibly undesirable, ways.

Anyway, some more specifics, and maybe a posted image or two, or even a posted project file, will probably go a long way to helping communicate the nature of the problem.

Rick_Gerard@adobeforums.com

2006-12-06, 8:15 pm

The problem is coming from not understanding the rendering pipeline. If the video is in a pre-comp or nested comp you need to collapse transformations so that scaling up the nested layer retains resolution. It sounds like you nested the video in a comp wi
th the note then animated the camera for each note, then added that comp to a main comp w/o a camera.

Not exactly the most efficient workflow, but it would work.

Here's what I would have done. Pre-composed each note with the video. Collapsed transformations. Animated the note, not the camera.

Mark_Silvestri@adobeforums.com

2006-12-06, 8:15 pm

Thanks, Aaron. I think you are right. It would be best for me to provide an AEP. I have cobbled one together complete with footage (minimum). However, I don't have a website or ftpsite set up to facilitate this. Is this something I can zip up and attach i
n an email reply on this forum? I didn't think attachments were allowed. What do you suggest?

Aaron_Cobb@adobeforums.com

2006-12-06, 8:15 pm

How big is the file? If it's a reasonable size, you can email it to me at aaron.s.cobb @ mac.com (be sure to remove the spaces before emailing), and I'll take a look at it when I have a moment.

Mark_Silvestri@adobeforums.com

2006-12-06, 8:15 pm

Thanks, Aaron, for agreeing to take a look.

However, again, I seem to have solved my problem. In preparing this AEP, while turning off motion blur, etc., I turned off the depth of field on my camera in the original note lift comp and voila! Despite your suggestion early on, I was convinced that it
was the scaling and not the depth of field that was causing the blurriness because I was able to maintain the video resolution - with depth of field on - using my clunky workaround as described earlier in this thread.

Now, I realize that the depth of field when the camera is 10x further away from the object (as in my clunky workaround comp) does not have nearly that same blurring effect as it does when the camera is upon the object (as in my original note lift comp). I
'm glad I caught this before you spent any time on it ... especially after you had pointed this out as a potential cause of my problem, earlier.

Thanks, again, Aaron and Rick.

Aaron_Cobb@adobeforums.com

2006-12-06, 8:15 pm

You can set the focus distance and the aperture to determine the depth of field. Narrowing the aperture will increase the DOF (i.e. the range of depths that are in focus) just as it would with a real camera.

Of course, working with one camera, per Rick's suggestion, will make this a LOT easier to deal with.